What do economists do? PhD job openings 2009-2013.

This post breaks down academic job openings for economists over the last five years by field. As reflected by job openings, what do economists study, and which fields have been “hot” over this time window?

My motivation for digging into these data is in part the response to my list of 18 symptoms of bad economics criticism. Paul Krugman and John Quiggin claimed that macro actually does dominate economics, although it’s not clear either of them actually disagrees with symptom #1 of bad criticism—the notion that economics as discipline is useless if we can’t forecast future states of the macroeconomy.

It was not my intention to start an interdisciplinary pissing match—macro is indeed a very important field, arguably the most important field in terms of policy relevance. But macro is just one field among many. Reflecting on the relative importance of macroeconomics made me wonder what proportion of economists are macroeconomists. And for that matter, how does the rest of the discipline break down by field? I have strong anecdotal impressions, but I don’t recall ever seeing statistics from representative data.

The market for economics PhDs is centralized. Almost all job openings are advertised in Job Openings for Economists, put out by the American Economic Association. The following results are based on the last five October issues, focusing on the desired fields (by standardized JEL codes) specified in the advertisement. (Employers often specify multiple fields in one ad, which I have handled by weighting such ads by the reciprocal of the number of fields listed. For example, an ad which lists the three fields Public, Labor, and Health is treated as 1/3 of a Public ad, 1/3 of a Labor ad, and 1/3 of a Health ad.)

Only a small proportion of jobs are in macro.

Macro is smaller than I would have guessed. The proportion of job openings for macroeconomists is small, as illustrated in the graph below. joe-graph0 Macro only forms about 10% of the market. This estimate may be attenuated somewhat because of the overlap between macro and other fields (notably International), but that issue cuts both ways.

Job openings by field.

As shown in the graph below, International is the most demanded field followed by Micro, “Mathematics and quantitative methods” (largely theoretical econometrics) and Macro. Moving down the list, we find a bunch of fields clustered together at about 50 openings give or take a dozen: Macro, Finance, Industrial Organization, Agricultural, Labor, Public, Development, and HEW (Health, Education, and Welfare, which is largely Health). Then remaining fields are small, added together they are comparable to one of the larger fields: “General,” Urban, Business Economics, Law and Economics, Economic History, History of Thought (which includes heterodox economics), and some miscellaneous fields with a just one or two jobs over the last five years.
joe-graph1

Short term trends.

PhD candidates will be happy to learn that the number of job openings has risen rapidly over the last five years, from about 200 to almost 350. Which fields are hot?
joe-graph2
The graph shows number of openings by field in 2009 and an arrow pointing to the number of jobs in the same field in 2013. International, Agricultural, Development, Labor, Public, and HEW experienced substantial increases in openings, whereas openings in Micro fall modestly and most other fields are fairly stable.

Some simple regression analysis largely confirms what we see in the graph. For each field, I regressed job openings on a linear trend. Along with openings per year and total openings, the estimated trend (\(\hat\beta\)) and the associated p-value against the null that there is no trend are displayed in the table below (these are regressions with n=5 and the p-values are based on OLS standard errors, so take with whatever size grain of salt you view as appropriate). Labor, Public, Finance, Agricultural, and International have expanded fastest, and all of these increases are statistically significant.

My code is a mess so I won’t post a link here, but email me if you would like me to send you a copy to play with.

Edit: Fixed an error with the second graph, updated descriptive text in preceding paragraph. October 30, 2013.

Field 2009 2010 2011 2012 2013 total \(\hat\beta\) p-value
International 12.4 14.7 17.6 16.8 18.7 80.1 1.5 0.02
Micro 13.0 14.3 10.9 15.9 9.5 63.6 -0.5 0.60
Quant/math 8.8 13.0 12.4 13.3 12.2 59.7 0.7 0.27
Macro 9.4 15.3 7.7 13.8 10.2 56.4 -0.0 0.99
AgEcon 7.2 10.3 10.2 14.0 14.2 55.9 1.8 0.01
Development 6.6 9.8 10.9 11.1 13.3 51.7 1.5 0.01
IO 8.0 8.4 7.5 8.5 11.2 43.6 0.7 0.17
Public 5.3 7.9 8.1 10.7 10.3 42.4 1.3 0.02
Health Ed Wfare 5.0 7.4 7.7 12.0 9.4 41.4 1.3 0.09
Finance 6.1 7.8 7.7 8.3 10.1 40.0 0.9 0.02
Labor 4.6 7.0 6.6 8.6 12.1 39.0 1.7 0.02
General 3.4 6.1 5.6 7.4 2.5 25.0 -0.1 0.95
All/default 5.5 3.5 5.8 3.7 3.3 21.8 -0.4 0.33
Business 3.0 2.3 3.3 1.9 4.7 15.1 0.3 0.46
Urban 2.6 2.3 2.8 3.7 3.4 14.8 0.3 0.09
History 2.1 1.5 2.0 1.9 2.5 10.0 0.1 0.36
Law 0.8 1.8 2.5 2.5 2.3 10.0 0.4 0.10
Hist thought 1.0 1.6 1.9 1.7 2.0 8.2 0.2 0.09
special 0.8 0.6 1.0 0.5 0.7 3.6 -0.0 0.70
systems 1.3 0.3 0.1 0.4 1.0 3.2 -0.0 0.82
misc 0.1 0.5 0.7 1.3 0.2 0.25

Tags: , ,

One Pingback/Trackback

  • Joel

    So few tenure track jobs in my field. Makes me feel less bad about not landing one back in2009.

    • Chris Auld

      I should have made it clearer: these are all academic postings, tenure track or not (although most of them are tenure track). I wish the data went back further so we could see the effect of the crisis, but if they exist in a convenient format I couldn’t find them in the several long seconds I spent searching, but I think 2009 was the bottom of the trench.

  • Evan

    How did you classify the “any field” searches? The second figure has an “all/default” label, but I don’t see it in the first.

    • Chris Auld

      Good catch, thanks. I just treated “all fields” as any other category, but the first (relevant) graph was inadvertently an old version which omitted “all fields” but included non-academic jobs. I’ve updated the graph and modified the text in the paragraph describing it.

  • Pingback: Somewhere else, part 87 | Freakonometrics

  • http://chrisauld.com/ Chris Auld

    “International” covers everything in JEL code F, “International Economics.” I do note in the post that many of these jobs may also be “macro,” but your point is well-taken.

    I did do all of the analysis above at the 2-digit JEL code level, too, but didn’t report it. At the 2-digit level the breakdown for code F looks like (sorry for the formatting)

    code jobs desc

    F 158 (all International)
    F1 49 Trade
    F2 6 International factor movements and business
    F3 43 International finance
    F4 43 Macroeconomic aspects of trade
    F5 10 International relations and political economy
    F6 6 Globalization

    So roughly half of the openings just list the 10digit code, F. Of those that do specify a 2-digit code, about one-third are (narrowly) macro.

    I agree generally that these classifications are quite nebulous, but I would still offer that the breakdown is interesting.

  • Dave Hedengren

    What method did you use to scrap the data? Cawley’s AEA paper on the job market says there were 1,300+ academic positions in 2010.

    • http://chrisauld.com/ Chris Auld

      Hi Dave, no scraping required—follow the JOE link above, click on one of the issues, then on the page that comes up click on “download and display options” and it’ll let you download the listings in spreadsheet form. John has more listings because he used all listings per year whereas I only used October listings.

Copyright © 2014 M. Christopher Auld